Department of Computer Science and Electrical Engineering

If your research involved language processing and/or machine learning you should consider submitting a poster abstract to the Mid-Atlantic Student Colloquium on Speech, Language and Learning.

This is a free, one-day event bringing together faculty, researchers and students from universities in the Mid-Atlantic area working in human language technology and/or machine learning. It is an opportunity for students to present preliminary or completed work and to network with other students, faculty and researchers working in related fields. The event will be held at JHU in Baltimore on Friday 23 September 2011.

Students are encouraged to submit one-page abstracts by Monday, August 15 describing ongoing, planned, or completed research projects, including previously published results and negative results. Submissions and presentations must be made by students or postdocs. See the call for papers for more information.

Accepted submissions will be presented as posters and each will also be given a one-minute presentation during a poster spotlight session. A small number of submissions will be selected to be presented as talks, on the basis of diversity and general interest. Student-led breakout sessions of one hour will also be held to discuss papers on topics of interest and stimulate interaction and discussion.

MS thesis defense

A Fast Fullchip Transient Response Estimation Technique

Sushmita K. Rao

11:00am Monday, 25 July 2011, ITE 346

Circuit Simulation has long been a dependable technique for design engineers for functional testing before a circuit is taken to silicon. But as we move into very deep sub-micron technologies, chips are becoming more complex and denser. The dense power grid adds to long simulation run time rendering fullchip simulations difficult and computationally expensive for larger designs. Some large designs may not be simulatable owing to system requirements. Even fast simulators like Cadence UltraSim fail to provide SPICE level accuracy. Dynamic solutions provide greater accuracy than statistical techniques but long simulation run time becomes the biggest tradeoff. Accurate measurement of dynamic currents is required for applications like power estimation, supply noise predictions and in verifying the power grid designs and testing. The research reported in this thesis is focussed on providing a current based method to estimate a fullchip's transient response without carrying out a fullchip simulation. Instead, a part of the chip is simulated and the power port currents thus measured are used to estimate the fullchip currents using the principle of superposition. This technique eliminates fullchip simulations that are complex and time consuming but the accuracy is maintained as the fullchip estimation is carried out on simulated data. A simple scaling method is also proposed to compensate for resistive variations in the test circuits. Experimental results reported demonstrate the effectiveness of the method.

Committee members:

• Dr. Chintan Patel (Chair)
• Dr. Ryan Robucci
• Dr. Tinoosh Mohsenin

The job search aggregator indeed.com collects job listings from a large number of of websites, including job boards, newspapers, associations, and company career pages. Using this data, they regularly report on job trends by country, region and keywords. Their most recent report by industry shows the weakness of the economic recovery but also shows a strong market for IT jobs.

Indeed has a web page that lets you see trends in the number of job postings with certain keywords. Amazingly, the current top ten trending phrases are all IT related: HTML5, mobile app, Android, Twitter, jQuery, Facebook, social media, iPhone, cloud computing and virtualization.

Mobile app Job Trends

Mobile App jobs

If you are preparing to look for a job or just trying to better prepare for a career, it is an interesting source of data. One thing to keep in mind is that you should invest your time in college building a foundation of knowledge and skills that will support you for the rest of your life, both professionally and as a well-rounded person. The undergraduate programs at UMBC aim to do just that. But you might give some thought to choosing some electives that will prepare you for the opportunities you will find in the short term, too. As usual, finding the right balance is the key.

Amazon has announced a new program allowing students to rent textbooks.

“Kindle Textbook Rental is a flexible and affordable way to read textbooks. You can rent for the minimum length, typically 30 days, and save up to 80% off the print list price. If you find you need your textbook longer, you can extend your rental by as little as 1 day as many times as you want and just pay for the added days.”

Amazon’s typically sells its kindle textbooks for nearly the same as the hardcopy ones, so renting a textbook may be attractive. We’ve not yet seen any Computer Science or Computer Engineering kindle books for rent, so can’t say how much of a discount there is if you rent for a semester.

By the way, the format that Kindle uses is based on Mobipocket, which is based on HTML. There are some good open source systems that you can use to create kindle-compatible documents from various sources.

While you can find a treehouse to rent using Airbnb, they also have a mockup of one for their employees to chill in.

Last week’s Wall Street Journal had a story, The Perk Bubble Is Growing as Tech Booms Again, on how technology companies compete for employees by offering more and more extravagant perks.

“Here in the capital of the latest tech boom, engineers and product developers work late into the night creating the next big thing. But they take office culture just as seriously, fueling behavior that is reaching a level of froth not seen in a decade.

Some Web start-ups are partying like it’s 1999. Airbnb’s housewarming later this month is to include a visit by rapper and occasional tech investor M.C. Hammer. The party room at reviews site Yelp Inc. has three beer kegs with built-in iPads to offer information about what’s on tap. Last month, start-ups Peanut Labs Inc. and AdParlor Inc. sponsored the sold-out “Pirates of Silicon Valley Cruise,” a $600-per-person seafaring party.

Ph.D. Dissertation Defense

Finding a Temporal Frame Comparison Function
for the Metacognitive Loop

Dean Earl Wright III

9:00am Wednesday 26 July 17 August 2011, ITE325 UMBC

The field of Artificial Intelligence has seen steady advances in cognitive systems. However, many of these systems perform poorly when faced with situations outside of their training. And as the real world is dynamic, this brittleness is a major problem in the field today. Adding metacognition to such systems can improve their operation in the face of perturbations found in dynamic environments. The Metacognitive Loop (MCL) works with a host system, monitoring its sensors and expectations. When a failure is indicated, MCL advises the host system on corrective actions.

Differing amounts of metacognition can be made available to the host system. At the lowest level no assistance is given. Above that are rule-based systems with hard-coded responses to stimulations. Next are evaluative systems that weigh multiple inputs using neural network or other techniques. At each level, the metacognition system can provide useful assistance in more situations.

The next level of metacognition adds a temporal dimension. A metacognitive system that has no concept of time has to either treat each new problem as (1) a symptom of previously encountered problems or (2) a completely new problem. Both of these extremes lead to providing poor advice to the host system. Thus, the temporal level of metacognition needs a more discriminating test to compare the current situation encountered by the host system with previous problems.

Several algorithms were developed to find the one that would provide the best performance for a simulated Mars Rover that faced with a dynamic environment with multiple perturbations. Several methods were found that allow MCL to provide good advice most of the time and allow the Mars Rover to successfully complete complex tasks.

Committee:

• Dr. Tim Oates (chair)
• Dr. Marie desJardins
• Dr. Tim Finin
• Dr. Anupam Joshi
• Dr. Don Perlis (UMCP)

Software Development Times reports that Computer Science is the top major for job offers.

“Computer science graduates now get more offers of employment than any other major. This is the first time since 2008 that computer science has topped the list: previously, accounting majors had the highest offer rate. In 2011, 56.2% of computer science majors received job offers, compared to only 53.8% of accounting majors. The offer rate for computer science majors increased 13.8% this year from the previous year.”

MS Thesis Defense

Distributed Approach for Mitigating Coverage Loss
in Heterogeneous Wireless Sensor networks

Kavin Rathinam Kasinathan

10:00am 15 July 2011, ITE 325b

In a heterogeneous wireless sensor network, nodes with different sensing capabilities are dispersed throughout an area of interest. Nodes with similar capabilities are not necessarily collocated. When a node fails, the area in the vicinity of such a node is left uncovered and the application may be negatively impacted. In this paper, we present a distributed algorithm for mitigating the coverage loss caused by the failure of a sensor node. The proposed algorithm looks for one or multiple nodes that can be repositioned in order to fill the coverage gap. A search is conducted within the 2-hop neighborhood to identify a node or a combination of more than one node that can collectively possess the capabilities of the failed node. In order to maximize the performance of the network, the replaceable nodes are chosen such that there is a minimum coverage and connectivity impact on the network after replacement. In the addition, the distance a node will travel to replace the failed node is also taken into consideration to limit the recovery overhead.

Committee members:

• Dr.Mohamed Younis (Chair)
• Dr.Charles Nicholas
• Dr.Tinoosh Mohsenin

CIO magazine has an article that identifies what they think are the the six hottest new IT jobs. They used an admittedly unscientific method of reviewing listing on IT job sites and talking to IT executives to find the types of jobs with good growth potential and are resistant to outsourcing and economic downturns.

“IT job seekers have real reason to hope. No fewer than 10,000 IT jobs were added to payrolls in May alone, according to the Bureau of Labor statistics, reflecting a steady month-over-month increase since January. And in a June survey by the IT jobs site Dice.com, 65 percent of hiring managers and recruiters said they will hire more tech professionals in the second half of 2011 than in the previous six months.”

Their six are:

• Business architect
• Data scientist
• Social media architect
• Mobile technology expert
• Enterprise mobile developer
• Cloud architect

While won’t find required courses on most of these in a standard undergraduate program, doing well in any of them needs the foundation you will receive. These include programming, software engineering, statistics, systems, computer architecture, algorithms, databases, etc. UMBC does offer electives that give students the skills that will make students more competitive for these jobs, such as mobile computing, parallel programming, service oriented architectures, machine learning, data mining, security, web technology, etc.

Disclaimer: Past performance does not guarantee future results. Your mileage may vary.

MS Thesis Defense

Clinical-Genomic Analysis for Disease Prediction

Darshana Dalvi

10:00am 6 July 2011, ITE 346

Recent advances in genomic research have generated vast amounts of information that can help identify individuals who differ in their susceptibility to a particular disease or response to a specific treatment. This information may offer solutions for the treatment of complex chronic diseases that are influenced by a wide array of factors. This vast amount of information brings critical challenges in applying advanced technology to synthesize clinical-genomic patient data. Synthesizing this information is necessary to derive the knowledge that would empower physicians to provide personalized care with the best possible therapeutic interventions.

We used statistical methods and data mining approaches to understand clinical-genomic risk factors that differentiate Type II Diabetes cases from healthy controls. We investigated whether inclusion of genomic risk factors in conjunction with clinical information improves classification accuracy. We also demonstrate how a biased and an unbiased method for selection of risk associated single nucleotide polymorphisms (SNPs) effect clustering along with clinical information. We determined the optimal method based on its clustering performance.

Committee members:

• Dr. Yelena Yesha (Chair)
• Dr. Michael Grasso
• Dr. Yaacov Yesha
• Dr. Milton Halem